Clinical predictors for bradycardia and supraventricular tachycardia necessitating therapy in patients with unexplained syncope monitored by insertable cardiac monitor

Abstract Background Insertable cardiac monitors (ICMs) improve diagnostic yield in patients with unexplained syncope. The most of cardiac syncope is arrhythmic causes include paroxysmal bradycardia and supraventricular tachycardia (SVT) in patients with unexplained syncope receiving ICM. Predictors for bradycardia and SVT that necessitate therapy in patients with unexplained syncope are not well known. Hypothesis This study aimed to investigate predictors of bradycardia and SVT necessitating therapy in patients with unexplained syncope receiving ICMs. Methods We retrospectively reviewed medical records of consecutive patients who received ICMs to monitor unexplained syncope. We performed Cox's stepwise logistic regression analysis to identify significant independent predictors for bradycardia and SVT. Results One hundred thirty‐two patients received ICMs to monitor unexplained syncope. During the 17‐month follow‐up period, 19 patients (14%) needed pacemaker therapy for bradycardia; 8 patients (6%) received catheter ablation for SVT. The total estimated diagnostic rates were 34% and 48% at 1 and 2 years, respectively. Stepwise logistic regression analysis indicated that syncope during effort (odds ratio [OR] = 3.41; 95% confidence interval [CI], 1.21 to 9.6; p = .02) was an independent predictor for bradycardia. Palpitation before syncope (OR = 9.46; 95% CI, 1.78 to 50.10; p = .008) and history of atrial fibrillation (OR = 10.1; 95% CI, 1.96 to 52.45; p = .006) were identified as significant independent predictors for SVT. Conclusion Syncope during effort, and palpitations or history of atrial fibrillation were independent predictors for bradycardia and for SVT. ICMs are useful devices for diagnosing unexplained syncope.


| INTRODUCTION
Syncope has various causes, and the prognosis differs according to the cause. 1 When conventional tests do not indicate the cause, the diagnosis is unexplained syncope. Of all patients with syncope patients in dedicated facilities, 18% to 20% had unexplained syncope. [2][3][4] Our previous report also indicated that there was 23.9% incidence of all syncope. 5 Insertable cardiac monitors (ICMs) allow for lengthy monitoring of cardiac rhythm and improved diagnostic yield among patients with unexplained syncope. [6][7][8] Remote monitoring systems with ICMs have also become available. With ICMs, physicians can intensively monitor patients with unexplained syncope. In most cases, cardiac syncope is arrhythmic. Sick sinus syndrome, atrioventricular block, and paroxysmal supraventricular tachycardia (SVT) have been found in patients with unexplained syncope that is monitored by ICMs. 9 When diagnosed through the use of ICMs, bradycardia, supraventricular tachycardia and ventricular tachycardia must be treated aggressively. In Western countries, patients who have unexplained syncope have been reported to have predictors for bradycardia that necessitates placement of a pacemaker 10-12 ; this phenomenon has not been reported in Asian countries. In addition, the predictors for SVT have not been reported. If these predictors can be clarified, clinicians could provide more specific targeted monitoring. We therefore aimed to identify these predictors.

| METHODS
We retrospectively reviewed medical records of consecutive patients who received ICMs to monitor unexplained syncope in three hospital facilities in Japan (Showa University Hospital,

| Statistical analysis
Data are reported as means ± standard deviations. Continuous and categorical variables were compared by means of the Mann-Whitney U test or chi-squared test, as appropriate. We performed Cox's stepwise logistic regression analysis to identify significant independent predictors that occurred during syncope and cardiovascular events and that were prognostic for bradycardia and SVT, and we calculated the odds ratios (ORs) and 95% confidence intervals (CIs) are presented. Using Kaplan-Meier curves, we analyzed the diagnosis on the day from ICM placed. We considered p values of less than .05 to be statistically significant. JMP software version 14.0 (SAS, Cary, NC, USA) was used for the analysis.

| RESULTS
We reviewed the medical records of 140 consecutive patients, of whom six were excluded because they changed hospitals during the study period. Two more patients were excluded because their ICMs had to be removed prematurely as a result of infection or skin erosion.

Abnormal
Because the ventricular tachycardia was lethal arrhythmia, the 2 patients with ventricular tachycardia received implantable cardiac defibrillators. These arrhythmias had induced syncope and, fortunately, had terminated spontaneously. Ventricular tachycardia took the form of concealed long QT in 1 patient (Figure 3(C)) and was caused by idiopathic ventricular fibrillation in 1 patient.
Syncope during effort (in 58%; p < .05) was significantly more frequent in patients with bradycardia who needed a pacemaker (Table 1 T A B L E 1 (Continued)

| DISCUSSION
We showed that syncope during effort, and palpitation, atrial fibrillation were independent predictors for bradycardia and for SVT. The main causes of syncope are known to be bradycardia, tachycardia (including SVT and ventricular tachycardia), noncardiac causes (including reflex syncope and orthostatic hypotension). 9,15 The arrhythmias must be aggressively treated because they affect quality of life and increase the risk of mortality.
ICMs are useful devices for diagnosing unexplained syncope. The annual cumulative diagnostic rate was calculated to be 43% to 50% over a maximum follow-up period of 2 years in previous reports in Western populations. [6][7][8] In those reports, the diagnostic rate increased rapidly during the 6 months period after ICM placement and linearly during the subsequent period.
Predictors of bradycardia have been previously reported in Western nations 10,11 but not in Japan. In a previous study, Ahmed et al reported that independent predictors for bradycardia that necessitated pacemaker implantation were female gender, age of more than 75 years, PR interval longer than 200 ms, and injury during syncope. 11 In another study, Palmisano et al reported that age of more than 75 years, injury during syncope, and bradycardia on ECG were independent predictors for bradycardia that necessitated pacemaker implantation. 10 Moreover, in both studies, the investigators reported that the presence of multiple predictors significantly increased the possibility that affected patients would need pacemaker implantation.
Those results, however, differed from ours. Some reasons are that Japanese and Western clinical settings may be different; the definition unexplained syncope may differ; and there may be physiological differences among various races. In addition, fewer Japanese patients with unexplained syncope may agree to have ICMs placed. 5 However, because syncope during effort has been reported to be a predictor for suspected cardiac syncope, 16 we believe that our finding of syncope as a predictor of bradycardias necessitating pacemaker implantation is related.
To the best of our knowledge, no previous reports have elucidated the predictors of SVT in patients with unexplained syncope monitored by ICM. SVT can lead to reduced cardiac output and syncope because of increased ventricular rate. In most cases of SVT, however, the heart rate is not rapid enough to impair ventricular function and cardiac output. 17 In a previous study, 20% of patients with SVT had at least episode of syncope which is preceded by palpitations. Multivariate analysis showed that heart rate ≥ 170 beats/min was the only independent predictor for syncope. 18 The mean rate of SVT induced syncope was 194 bpm, most of patients was above 170 bpm in present study. The reason is considered that none of registered patients have severe structural heart disease such as low ejection fraction and cardiomyopathy. Syncope symptom is generally rare in patients with SVT; symptoms mostly reflect palpitation. 18 We also  12 In an earlier report, 0% to 13% of patients with unexplained syncope monitored by ICM had ventricular tachycardia. 10,24,25 Although no patients died during observation period in our study, a minority of patients with unexplained syncope monitored by ICM did die in previous studies. 11,12 Clinicians should remember that patients with unexplained syncope include those with potentially lethal ventricular tachycardia or other risks for mortality.

| STUDY LIMITATIONS
This study was observational and retrospective, and the findings need to be confirmed in a larger and longer trial. In addition, our data appear to have a selection bias. This is dependent on physicians who diagnose unexplained syncope and on interpretation of ICM data.
Indeed, we may have overidentified predictors because a few patients had bradycardia or SVT that necessitated therapy.

| CONCLUSIONS
ICMs are useful devices for diagnosing unexplained syncope. Syncope during effort, and palpitations or history of atrial fibrillation were independent predictors for bradycardia and for SVT. We should carefully follow up of patients with these predictors.